Diffuse large B-cell lymphoma (DLBCL) is the most common non-Hodgkin lymphoma (NHL) subtype. It is well established that there are driver somatic mutations in DLBCL etiology and prognosis, as well as a role for germline genetic susceptibility. In fact, early results from candidate gene association studies have shown a promising role for common genetic variants in immune and apoptotic genes in NHL. In the last funding cycle, we identified SNPs in BCL2L11 (BIM), CASP9, and APAF1 that were associated with increased risk in our SPORE case-control study. Resequencing of these genes in the tumors of 40 DLBCL patients identified novel genomic alterations, and in Aim 1 we propose to characterize the etiologic and therapeutic significance of these novel mutations with laboratory-based studies. In Aim 2 and 3, we propose a comprehensive and agnostic strategy that integrates both somatic and germline genetics in order to identify additional novel risk variants. We will do this by leveraging our SPORE's involvement in the large International Lymphoma Epidemiology (InterLymph) consortium genome-wide association study (GWAS) of DLBCL (>S000 cases and 10,000 controls) and our whole-exome next generation sequencing (NGS) study of paired tumor and germline DLBCL cases (N=77). The GWAS is powered to identify common variants, and the NGS study will allow us to identify lower frequency variants, defined here as 0.5% to 5%. Using a multistage design in Aim 2, we propose to identify novel gennline low-frequency variants associated with risk of developing DLBCL, and in Aim 3, we identify and validate somatically acquired driver mutations in genes critical to DLBCL pathogenesis. In exploratory analyses, we will evaluate pathways and the role of these variants in DLBCL prognosis. This proposal utilizes the unique resources of our SPORE, the InterLymph GWAS, and our DLBCL whole-exome NGS project. We have an outstanding team with a strong track record of interdisciplinary genetics work in lymphoma and have demonstrated the ability to integrate genetic epidemiology with lab-based functional work. Our study builds on several novel findings from the prior study period, but also expands to fill an important need in the genetic epidemiology of DLBCL as the first comprehensive study of low frequency germline variants in risk. Low frequency variants, either individually or cumulatively across a gene, and in combination with common variants, are likely to inform etiologic pathways and clinical risk assessment. Furthermore, we will identify novel driver mutations in genes and pathways from DLBCL tumors that can inform tumor biology and identify novel treatment targets. In summary, as the first comprehensive study of both germline and somatic genetic variants in DLBCL, we are likely to provide new and unexpected insights into lymphomagenesis, which can then be exploited clinically for risk assessment, prognostic stratification and identification of new treatment targets.